I built a Social Media Performance Coach
The AI coach keeping my content goals on track, every single week.
My building spree continues this week with an automated performance report for my socials. I used Claude Cowork and programmed it to run every Monday at 7 am. It audits LinkedIn, Instagram, TikTok and Youtube and gives me performance metrics and content recommendations. It’s like my own social media performance coach.
WHAT COWORK IS
Claude Cowork is Anthropic’s desktop agent. It doesn’t just answer questions, it takes action on your computer. It can open apps, navigate websites, pull data, and compile it into whatever format you ask for.
For creators and marketers, that means you can give it a task like “analyze my social media performance this week” and it will actually go do it across every platform, in one session.
THE PROMPT
Here’s the exact prompt I use. Copy it, swap in your own profile links, and run it in Cowork:
Can you do an analysis of everything I posted last week and what performed the best? What should I do less of? Can you compile it in HTML with graphs and callouts. LinkedIn link: [your LinkedIn profile URL] Instagram link: [your Instagram profile URL] TikTok link: [your TikTok profile URL] YouTube link: [your YouTube channel URL]
✨ Cowork will open each platform, pull your recent posts, analyze performance, and output a formatted HTML report you can save and reference all week.
WHAT MY REPORT SHOWED THIS WEEK
This week’s report surfaced three things I wouldn’t have caught manually:
1. TikTok is my fastest growing platform right now — +212% views week over week. And 13.8% of those views came from Search, meaning people who don’t follow me are finding my content. Granted, I did boost some posts but it was cool to see some traction. That’s the signal I needed to double down on TikTok.
2. My best LinkedIn posts weren’t the obvious ones. The post that hit 664 impressions wasn’t about a major AI headline, it was a niche infrastructure angle nobody else was framing. Non-obvious takes win.
3. I was batch-posting on LinkedIn and it was killing my reach. Eight posts in one morning averaged 12–35 impressions each. Spacing them out is the fix.
TRY IT THIS WEEK
Swap in your links, run the prompt in Cowork, and reply to this email and tell me what your report shows.
If you’re not on Cowork yet, it’s a desktop app from Anthropic. Worth the download just for this use case.
📰 Top Stories This Week
AI & TOOLS
Reddit CEO Declares User Content the ‘Modern Oil’ for AI
Reddit CEO Steve Huffman says large language models wouldn’t exist without the user-generated content on platforms like Reddit and he’s been making deals with Google and OpenAI to prove it.
💡 Why this matters This is a reminder of just how valuable your content really is. Every post, comment, and thread you publish is contributing to the next generation of AI. For creators, that opens up a real conversation about data ownership and what fair compensation might look like down the road.
CREATOR ECONOMY
The Digital Media Reckoning: What Happened to the Darlings of the 2000s?
BuzzFeed, Vox Media, and other digital empires that were once valued in the billions are now struggling selling off assets and scrambling to find a new model that works.
💡 Why this matters This is a cautionary tale for every creator and marketer. The platforms that looked unstoppable a decade ago didn’t survive the shift. Diversify your revenue streams, don’t bet everything on one platform, and always be stress-testing your business model.
AI & TOOLS
Google’s New ‘Anything-to-Anything’ AI Model Wows
Google unveiled a new AI model that can take any type of input — text, images, audio — and turn it into any type of output. Early looks suggest it could reshape how creative content gets made.
💡 Why this matters We’re moving into a world where ideas can be transformed across formats instantly. A script becomes a video. An image becomes a soundtrack. For creators, that’s a production upgrade. Start thinking about what stories you could tell if format was no longer a bottleneck.
AI & TOOLS
Scaling Creativity in the Age of AI
A new piece from MIT Technology Review makes the case that storytelling is still fundamentally human — and AI is just the latest medium for expressing and distributing it at scale.
💡 Why this matters This is the reframe you need. AI isn’t replacing creativity; it’s amplifying it. The creators who will win are the ones who figure out how to use them to tell bigger, better, more resonant stories.
🎯 Your Edge This Week
Since Reddit content is being used as high-value fuel for the next generation of AI, start repurposing your best newsletter insights or long-form guides into Reddit threads this week. You’ll build authority with a real human audience — and make sure your perspective is the one AI models learn from.
📊 This Week’s Big Patterns
Pattern 1: The Agentic Takeover: AI Is No Longer Waiting to Be Asked
Something shifted this week. AI is becoming something that runs in the background, taking action on your behalf without being prompted.
Google just redesigned its search engine to work more like an assistant that fulfills your request than a list of links to click. Advertising platforms are testing systems that plan and adjust campaigns on their own. And Alibaba’s new AI can work autonomously for over 35 hours straight.
The shift from “AI that helps” to “AI that does” is happening faster than most people realize.
📰 Supporting Stories
• Google Search Goes ‘Agentic,’ Minimizing User Input
• The Rise of Agentic Advertising: AI Taking the Reins?
• Alibaba’s Qwen3.7-Max: The Autonomous AI Running for Days
🔮 Prediction: Within 18 months, the #1 metric for online visibility won’t be how often people click your link. It’ll be how often an AI agent cites your brand when answering someone’s question. Direct website visits will become secondary.
❓ Ask yourself: If an AI can fulfill your customer’s request without them ever landing on your website, what does your digital presence actually need to offer?
Pattern 2: The Bifurcated Web: You’re Now Writing for Two Audiences
The internet is quietly splitting into two tracks. One is built for humans — emotional, visual, and designed to resonate. The other is built for AI agents — structured, factual, and easy for machines to read and cite.
Google is now embedding deepfake detection directly into search results. Reddit’s CEO is talking about user content as the raw material powering AI. And forward-thinking publishers like The Economist are already building separate experiences for human readers and AI crawlers.
The content that wins going forward will be legible to the systems that summarize the internet for people.
📰 Supporting Stories
• Reddit CEO Declares User Content the ‘Modern Oil’ for AI
• The Economist Prepares for a Dual Internet: Human and AI Agents
• Google Embeds AI Content Verification Directly Into Search
🔮 Prediction: We’ll see the rise of ‘Proof of Human’ labels for premium content where creators use verified metadata to prove their work was made without generative AI. That badge will command 2x–3x the price of AI-assisted content.
❓ Ask yourself: Are you building a brand that AI agents will trust enough to recommend or are you just producing more content for the machine to digest?
Pattern 3: The Lean Logic Shift: Fewer People, Bigger Output
The collapse of BuzzFeed-era media giants isn’t just a nostalgia story, it’s a warning. Big teams, big overhead, and rigid 10-year plans are getting wiped out by smaller, faster, system-driven operations.
Meta just cut 10% of its workforce. Sheryl Sandberg is publicly telling people to ditch long-term career plans. And research shows that most companies aren’t actually winning with AI. They’re just using it to do the same things they’ve always done, slightly faster.
The companies pulling ahead aren’t the ones cutting people. They’re the ones building systems that let a small team punch far above their weight.
📰 Supporting Stories
• The Digital Media Reckoning: What Happened to the Darlings of the 2000s?
• Sheryl Sandberg: Ditch the 10-Year Plan, Embrace Agility
• Are You Really Winning at AI, or Just Using It?
🔮 Prediction: The next media company to reach BuzzFeed-scale impact will have fewer than 10 people. A fleet of AI agents handling production, distribution, and ad operations will make 90% profit margins possible for the right team.
❓ Ask yourself: Is your AI strategy about doing the same work with fewer people — or doing 10x the work with the people you already have?
See you next week! — Karina



